Waste Conversion into a Sweetener—Development of an Innovative Strategy for Erythritol Production by Yarrowia lipolytica

: The study proposed the innovative low-cost strategy for erythritol production by Yarrowia lipolytica through developing a simple medium based on industrial waste by-products and a natural method for culture broth puriﬁcation. Results obtained proved that corn steep liquor might successfully replace traditional sources of nitrogen and other nutrients without compromising activities of the enzymes responsible for erythritol production and its production level. As a consequence, a production process was performed where Y. lipolytica A-6 was able to produce 108.0 g / L of erythritol, with a production rate of 1.04 g / Lh and a yield of 0.45 g / g of the medium containing exclusively 220 g / L of crude glycerol derived from biodiesel production and 40 g / L of corn steep liquor. Moreover, a comparable concentration of erythritol (108.1 g / L) was obtained when a part of crude glycerol was exchanged for the crude fraction of fatty acids in the two-steps process. Next, the collected post-fermentation broths were used in the culture with Y. lipolytica Wratislavia K1 for natural puriﬁcation. The process resulted in a high increase of erythritol selectivity from 72% to 97% and in the production of 22.0 g / L of biomass with 40.4% protein content, which enables its use as an attractive animal feedstu ﬀ . KA— α -ketoglutaric acid, CA—citric acid.


Introduction
In the last decade, modern trends in nutrition led to the increased consumption of erythritol. This polyol has many advantages, the primary of which includes low energy value (0.2 kcal/g), lack of bad aftertaste, not affecting insulin level in the blood, and very good tolerance in the human digestive system, which makes it one of the best sugar substitutes [1].
The market demand for erythritol has meant that, in recent years, many works have been devoted to it and reported the conditions of its biosynthesis [2]. The possibility of production cost reduction was investigated by, e.g., screening the best-producing strains and their genetic modifications [3,4], the use of various culture systems [5][6][7], optimization of culture conditions [8][9][10], and the use of various substrates [8,[11][12][13][14]. Among the various substrates tested, the use of the media based on industrial waste by-products seems to be the most prospective for biotechnological processes, by being both cost-effective and environmentally-friendly.
Yarrowia lipolytica yeast is one of the most extensively studied organisms in respect of erythritol production. The environmental conditions, as well as the biochemical pathways of this polyol biosynthesis, have been widely described [2,15,16]. This non-conventional yeast has low growth requirements, is able to assimilate a wide spectrum of substrates, and has fair resistance to impurities

Substrates
Carbon and energy sources in the media applied included: pure glycerol (98%), crude glycerol from methyl ester production (SG BODDINS GmbH; Germany) containing 86% wt/wt of glycerol, 6.5% wt/wt of NaCl, and methanol-below 0.2%, and the crude fraction of fatty acids (CFFA) remained after the ethanolic esterification of fatty acids contained in rapeseed oil (pilot installation at the University of Wroclaw; Poland).

Media
Two types of growth medium were used for seed culture preparation: glycerol-based growth medium (GGM) for the inoculation of mineral shake-flasks medium (MS-FM) and bioreactor media (mineral production medium-MPM; low cost production medium-LPM; purification process medium-PPM), and fatty acid crude fraction-based growth medium (FGM) to inoculate bioreactor cultures (fatty acid crude fraction based production medium-FPM and waste substrates based production medium-WPM) in which the production medium contained CFFA. The growth media (GGM, FGM) consisted of (g/L): CFFA, 20.0, or pure glycerol, 50.0, as the substrates (in FGM and GGM medium, respectively), yeast extract, 3.0, malt extract, 3.0, and bactopeptone, 5.0, dissolved in distilled water.
Erythritol biosynthesis in the shake-flasks experiment was performed in the mineral medium (MS-FM) that consisted of (g/L): Pure glycerol, 100.0, MgSO 4  tap water, pH 3.0. To obtain 24.0 or 40.0 g/L of CSL, 48.0 or 80.0 g/L of 50% CSL (Sigma-Aldrich) was introduced to the medium. To prepare the purification process medium (PPM), the culture broth after earlier performed erythritol biosynthesis processes was collected, separated from the biomass by filtration (ø 0.45 µm), sterilized (121 • C, 20 min), and stored at 2 • C. Finally, the previously prepared liquids were supplemented with (g/L): (NH 4 ) 2 SO 4 4.6, MgSO 4 × 7H 2 O, 1.0, KH 2 PO 4 , 0.25, and yeast extract, 1.0, pH 3.0/4.5. In the processes with FPM and WPM media, an appropriate glycerol portion was introduced to the culture after 24 h of cultivation, whereas, in the case of the MPM and LPM media, glycerol was present from the beginning of the process. All the media were sterilized at 121 • C for 20 min.

Culture Conditions
The growth cultures (GGM, FGM) were carried out in 0.3-L flasks containing 0.1 L of the growth medium on a rotary-shaker (CERTOMAT IS, Sartorius Stedim Biotech GmbH) at 29.5 • C and 140 rpm for 3 days. In order to inoculate the media for the shake-flask and bioreactor experiments, 3 mL and 200 mL of the growth culture were used, respectively. The shake-flasks experiment was conducted for 7 days in 0.3-L flasks containing 0.03 L of the appropriate production medium (MS-FM), inoculated with 1 mL of seed culture, in the same conditions as described above. The samples were taken for analysis at the end of the experiment. Bioreactor cultures were carried out in a 5-L stirred-tank reactor (Biostat B Plus, Sartorius, Germany) with a working volume of 2 L, at 30 • C, the aeration rate fixed at 0.6 v/v/min, and the stirrer speed adjusted to 800 rpm. The pH of 3.0 was maintained automatically by the addition of a 20% (w/v) NaOH solution. All the bioreactor cultures were cultivated untill glycerol was completely consumed. Samples were withdrawn 2-3 times per day. Results are presented as mean values of three biological replicates for the shake-flasks experiment and two biological replicates for the cultures conducted in bioreactor.

Analytical Methods
Biomass content of the samples was determined gravimetrically after drying at 105 • C. Concentrations of glycerol, erythritol, mannitol, arabitol, citric acid, and α-ketoglutaric acid were determined with the HPLC method [14]. The osmotic pressure was measured with a Marcel OS 3000 osmometer (Marcel, Poland).

Statystical Analysis
To detect significant differences in the data, one-way analysis of variance was performed using Statistica 12.5 software (StatSoft, Tulsa, OK, USA). Homogeneous groups were determined using the Duncan's test (p ≤ 0.05).

The Effect of Crude Fatty Acid Fraction Application
Three natural strains, A-3, A-6, and A-311, of Y. lipolytica were examined for erythritol biosynthesis in the two-stage process (Table 1) in the FPM medium. In the first, growth stage only, CFFA was present in the medium as the source of carbon and energy. Next, after 24 h of cultivation, the erythritol production phase was initiated by glycerol addition to the culture. All the strains under investigation were able to assimilate both substrates; however, the level of biomass in the stationary phase depended on the strain and was the highest for the A-6 strain (about 34.4 g/L). The process was run until complete depletion of glycerol, which lasted from 92.0 h in the case of strain A-6 to 169.5 h when strain A-311 was applied, resulting in erythritol production of 66.0 to 86.3 g/L, respectively. Simultaneously, by-production of mannitol exceeded 21.0 g/L for strain A-3 and 40.0 g/L for the two other strains under investigation. Additionally, some culture media were supplemented with NaCl in two variants: salt applied as a component of initial medium or by its simultaneous addition with glycerol portion after 24 h of the cultivation process. The use of NaCl increased the osmotic pressure of the culture media (from 0.2 up to 3.4 Osm/kg) and allowed for higher erythritol production (up to 125.3 g/L) and significant improvement of the production yield due to reduction of mannitol by-production. The best parameters of erythritol production, i.e., productivity of 1.19 g/Lh and specific production rate of 0.035 g/gh, were obtained for the strain A-6 in the culture in which NaCl was present in the medium from the beginning of the process. Conditions: fraction-based production medium (FPM) with CFFA 20 g/L and pure glycerol 200 g/L (addition after 24-h of cultivation). Means in the same column marked with different letters (a, b, . . . ) are significantly different; p ≤ 0.05. * Total time of the process; ** Calculated for glycerol conversion in respect of the time when it was introduced to the culture broth; *** Initial osmotic pressure of the culture broth and its change resulting from glycerol addition. X-biomass, ERY-erythritol, MAN-mannitol, Y ERY -erythritol production yield, Q ERY -volumetric erythritol production rate, q ERY -specific erythritol production rate.

The Effect of CSL Application
The impact of CLS on erythritol biosynthesis from glycerol by the A-6 strain of Y. lipolytica was examined in the shake-flask ( Figure 1) and bioreactor cultures (Table 2), respectively in the MS-FM and MPM media. In the shake-flask experiment, the increase of CSL concentration from 0.0 to 40 g/L resulted in the increase of biomass level from 4.3 to 10.5 g/L and erythritol concentration from 6.1 to 28.5 g/L ( Figure 1). With the enhanced erythritol biosynthesis, the volumetric and specific production rates, as well as the yield of erythritol production, improved up to 0.17 g/Lh, 0.016 g/gh, and 0.32 g/g, respectively. However, the strongest stimulatory effect was observed when the CLS concentration was increased to 24.0 g/L, whereas its further increase to 40.0 g/L had no longer such a strong impact on erythritol production. Therefore, both variants were chosen for comparison with the bioreactor Sustainability 2020, 12, 7122 5 of 13 production cultures performed in the MPM medium that contained 150 g/L of glycerol and 25 g/L of NaCl. The process with the higher dose of CSL was almost half as much (147 h), and the yeast biomass level was twice as much (25.4 g/L) in comparison to the culture with 24.0 g/L of CSL (Table 2). Moreover, the application of 40.0 g/L of CSL enhanced erythritol biosynthesis, causing erythritol production at 83.0 g/L with a yield of 0.56 g/g and productivity of 1.08 g/Lh.
Sustainability 2020, 12, x FOR PEER REVIEW 5 of 13 respectively. However, the strongest stimulatory effect was observed when the CLS concentration was increased to 24.0 g/L, whereas its further increase to 40.0 g/L had no longer such a strong impact on erythritol production. Therefore, both variants were chosen for comparison with the bioreactor production cultures performed in the MPM medium that contained 150 g/L of glycerol and 25 g/L of NaCl. The process with the higher dose of CSL was almost half as much (147 h), and the yeast biomass level was twice as much (25.4 g/L) in comparison to the culture with 24.0 g/L of CSL (Table 2). Moreover, the application of 40.0 g/L of CSL enhanced erythritol biosynthesis, causing erythritol production at 83.0 g/L with a yield of 0.56 g/g and productivity of 1.08 g/Lh.

Comparison of Factors Causing the Increase of the Osmotic Pressure
Biosynthesis of erythritol from glycerol by the A-6 strain of Y. lipolytica was compared in terms of an increased osmotic pressure caused by an increased substrate concentration or NaCl presence in the MPM medium with 40.0 of g/L CSL. In the case of both factors, the increase of the osmotic pressure resulted in the significant shortening of the process and increased erythrtiol production ( Table 3). The yield and productivity of erythritol production were improved; however, the values of these parameters were higher in the case of NaCl application and reached 0.49 g/g and 1.0 g/Lh, respectively. The analysis of enzyme activities showed that the higher dose of the substrate stimulated the activity of glycerol kinase ( Table 3). The stimulatory effect of glycerol and NaCl was similar in the case of transketolase activity, which reached 0.064 and 0.069 U/mg, respectively. However, it was noted that erythrose reductase activity (0.120 U/mg) was enhanced to a greater extent by salt presence in the medium. Table 3. Biosynthesis parameters and enzymatic activity during erythritol production by Y. lipolytica A-6 depending on the factor causing the increase of the osmotic pressure in the medium.

Comparison of Mineral and Low-Cost Medium
Pure and crude glycerol (220 g/L) were used in the mineral (MPM containing 40.0 g/L of CSL and 25.0 g/L of NaCl) and low-cost (LPM) media used for erythritol biosynthesis by Y. lipolytica A-6 ( Table 4). When pure substrate was used, about 93.0 g/L of erythritol was produced in both the mineral and the low-cost medium, which corresponded to the yield of about 0.40 g/g. Elimination of mineral components in the low-cost medium led to prolonged biosynthesis time (159 h) and a twice lower biomass level (16.4 g/L), which resulted also in the lower productivity and specific production rate, in comparison to the culture with the mineral medium. The use of crude glycerol resulted in very similar results of erythritol biosynthesis obtained in mineral and low-cost media. In both media, within about 100 h, yeast were able to produce about 110 g/L of erythritol with the productivity of about 1.1 g/Lh and specific production rate of 0.040 g/gh.

Development of Waste-Substrate Based Medium
The WPM medium, containing only natural and waste substrates (see Materials and Methods Section), was proposed based on the previous experiments. The process of erythritol biosynthesis with the A-6 strain of Y. lipolytica was performed in two stages, according to the strategy carried out earlier in the process with the FPM medium: Biomass was produced when only CFFA was present in the culture medium, and erythritol biosynthesis was initiated after about 24 h by pure or crude glycerol introduction to the culture broth. In the culture with 150 g/L of crude glycerol, biomass concentration reached 20.3 g/L and yeast were able to produce 63.5 g/L of erythritol (Table 5) with a yield of 0.42 g/g and a productivity of 0.42 g/Lh. The application of the higher concentration (200 g/L) of the substrate resulted in a process extension to 208.5 h; however, the biomass level and parameters of erythritol biosynthesis increased. In the culture with pure and crude glycerol (200 g/L), erythritol concentration reached 110.3 and 108.1 g/L, which corresponded to the productivity of 0.67 and 0.57 g/Lh, respectively.

Natural Purification Process
The purification process, which was intended to increase the selectivity of erythritol in the post-culture broth, was performed in the PPM medium at pH 3.0 and 4.5 with the use of Y. lipolytica Wratislavia K1 strain. The PPM medium initially contained 84.4 g/L of erythritol, 15.5 g/L of mannitol, 8.3 g/L of arabitol, and citric and ketoglutaric acids in the concentration of 2.6 and 2.9 g/L, respectively ( Figure 2). Hence, the selectivity was 72%. At pH 3.0, after 64 h cultivation, the selectivity of the broth increased to 99% and 14.6 g/L of the biomass with 38.2% of protein content being produced. The increase of the pH value to 4.5 resulted in the purification process shortening to 49 h. After the process, the selectivity reached 96%, whereas biomass concentration was 14.6 g/L and its protein content was 40.4%.

Discussion
The cost of the fermentation medium is a critical factor for the industrial biosynthesis process. Therefore, the cost-effectiveness of biosynthesis significantly depends on the cost of the substrate and the other components of the production medium. The natural feature of Y. lipolytica yeast is the capability for assimilation of hydrophilic and hydrophobic substrates, which allows the use of a wider range of substrates, including low-cost renewable resources or crude industrial wastes [21]. However, the biochemical pathways of utilization of these two types of substrates are different, which makes their use dependent on the specific direction of biosynthesis and the desired product to be obtained. In the first step of the presented study, two types of substrates were applied: glycerol and crude fatty acid fraction. Y. lipolytica yeast use the phosphorylation pathway, in which glycerol kinase (GK) catalyzes glycerol phosphorylation to glycerol-3-phosphate, which is subsequently converted via dehydrogenation by glycerol-3-phosphate dehydrogenase (GPDH) to dihydroxyacetone-phosphate that might be integrated into different pathways, resulting in biomass, polyols, lipids, citric, pyruvic, or alpha-ketoglutaric acids formation [27,28]. In turn, fatty acids are directly metabolized into acetyl-CoA throughout β-oxidation. Acetyl-CoA is next incorporated to the citric acid cycle and might be easily converted to different acids of the Krebs cycle or used for biomass growth [29,30]. Considering different biochemical routes of utilization, the process of erythritol biosynthesis (Table 1) was divided into two steps: the first stage of biomass growth on the CFFA as the sole substrate in the medium, and the second stage of erythritol formation induced by glycerol introduction to the culture broth (with simultaneous increase of osmotic pressure of the culture medium). Different abilities to utilize and to grow on the CFFA were observed for the strains under investigation, resulting in different levels of biomass produced. Erythritol production was accompanied by the synthesis of a relatively high amount of mannitol (21.9-43.0 g/L). Earlier studies have shown that the by-production of mannitol might be reduced by increasing the osmotic pressure of the medium [31,32]. Therefore, a higher osmotic pressure was induced by the addition of 25 g/L of NaCl in two variants: at the beginning and after 24-h of the cultivation process ( Table 1). The method of NaCl incorporation to the culture had no significant impact on biomass level but enhanced erythritol production to 125.3 g/L and reduced mannitol by-production to 75%. Yang et al. [32] reported that the increase of osmotic pressure to 3.89 Osm/kg in the cultures with Y. lipolytica grown on 200 g/L of glycerol, achieved by initial medium supplementation with 20 g/L of NaCl, decreased biomass production (from 16.0 to 11.9 g/L), increased erythritol production from 55.7 to 86.7 g/L, and

Discussion
The cost of the fermentation medium is a critical factor for the industrial biosynthesis process. Therefore, the cost-effectiveness of biosynthesis significantly depends on the cost of the substrate and the other components of the production medium. The natural feature of Y. lipolytica yeast is the capability for assimilation of hydrophilic and hydrophobic substrates, which allows the use of a wider range of substrates, including low-cost renewable resources or crude industrial wastes [21]. However, the biochemical pathways of utilization of these two types of substrates are different, which makes their use dependent on the specific direction of biosynthesis and the desired product to be obtained. In the first step of the presented study, two types of substrates were applied: glycerol and crude fatty acid fraction. Y. lipolytica yeast use the phosphorylation pathway, in which glycerol kinase (GK) catalyzes glycerol phosphorylation to glycerol-3-phosphate, which is subsequently converted via dehydrogenation by glycerol-3-phosphate dehydrogenase (GPDH) to dihydroxyacetone-phosphate that might be integrated into different pathways, resulting in biomass, polyols, lipids, citric, pyruvic, or alpha-ketoglutaric acids formation [27,28]. In turn, fatty acids are directly metabolized into acetyl-CoA throughout β-oxidation. Acetyl-CoA is next incorporated to the citric acid cycle and might be easily converted to different acids of the Krebs cycle or used for biomass growth [29,30]. Considering different biochemical routes of utilization, the process of erythritol biosynthesis (Table 1) was divided into two steps: the first stage of biomass growth on the CFFA as the sole substrate in the medium, and the second stage of erythritol formation induced by glycerol introduction to the culture broth (with simultaneous increase of osmotic pressure of the culture medium). Different abilities to utilize and to grow on the CFFA were observed for the strains under investigation, resulting in different levels of biomass produced. Erythritol production was accompanied by the synthesis of a relatively high amount of mannitol (21.9-43.0 g/L). Earlier studies have shown that the by-production of mannitol might be reduced by increasing the osmotic pressure of the medium [31,32]. Therefore, a higher osmotic pressure was induced by the addition of 25 g/L of NaCl in two variants: at the beginning and after 24-h of the cultivation process ( Table 1). The method of NaCl incorporation to the culture had no significant impact on biomass level but enhanced erythritol production to 125.3 g/L and reduced mannitol by-production to 75%. Yang et al. [32] reported that the increase of osmotic pressure to 3.89 Osm/kg in the cultures with Y. lipolytica grown on 200 g/L of glycerol, achieved by initial medium supplementation with 20 g/L of NaCl, decreased biomass production (from 16.0 to 11.9 g/L), increased erythritol production from 55.7 to 86.7 g/L, and decreased mannitol concentration from 50.4 to 27.7 g/L. In the present study, in the cultures with the rapid increase of osmotic pressure resulting from simultaneous introduction of NaCl and a high concentration of glycerol, the total time of the biosynthesis was prolonged, which is in agreement with earlier reports [31]. However, in the presented work, a very desirable effect of process shortening was obtained in the culture with NaCl, which improved also the erythritol productivity parameters in the cultures in which yeast cells adapted to the conditions of increased osmotic pressure from the beginning of the cultivation. The high osmotic pressure (2.76 Osm/L), induced by the addition of 80 g/L of NaCl, was reported as necessary for effective erythritol biosynthesis by Y. lipolytica from a waste cooking oil and allowed the production of 21.8 g/L of erythritol with the yield 0.80 g/g in the shake-flasks experiment and 22.1 g/L of the polyol with the yield of 0.74 g/g in the bioreactor culture [21,22].
Apart from the substrate, another necessary component of the medium is the nitrogen source. Yeast extract and peptone are usually used in biosynthesis processes as the organic sources of nitrogen; however, their use should be avoided due to their high cost. CSL was reported as an economic and effective substitute for the traditionally used nitrogen sources, e.g., in the production of bioemulsifier, biodiesel, erythritol, and citric and succinic acids by different microorganisms, including Y. lipolytica [7,21,[33][34][35]. This organic component is a by-product of the corn steeping process of the maize-starch industry, and is rich in amino acids, vitamins, and polypeptides. In the presented study, the effect of CSL concentration on erythritol production was investigated in shake-flask ( Figure 1) and bioreactor batch cultures ( Table 2). In the shake-flask experiment, the increasing concentration of CSL enhanced erythritol biosynthesis, but the impact of CSL concentrations above 24 g/L was very small. However, the process performed in the bioreactor proved that the application of 40.0 g/L of CSL resulted in a significant improvement, especially in the parameters of erythritol production, as the process shortened to 77 h, which was half in comparison to the process with 24.0 g/L of CSL. Generally, the enhanced biomass growth results in decreased formation of process metabolites. Therefore, noteworthy is the fact that, in the presented experiment ( Figure 1, Table 2), the increased concentration of CSL stimulated biomass growth; however, the yield of erythritol production increased simultaneously. The same observation was reported earlier for erythritol biosynthesis by Yarrowia lipolytica Wratislavia K1 [7] and was supposed to be the effect of vitamins and other component of the CSL, as it is a complex substance of natural origin. Under similar conditions (40.0 g/L of CSL and 150 g/L of glycerol), Wratislavia K1 strain was able to produce 61.0 g/L of erythritol with the productivity of 0.85 g/Lh and the yield of 0.38 g/g, which was significantly lower than in the case of the A-6 strain of Y. lipolytica in the present study.
NaCl is a functional component of the medium, added to increase the osmotic pressure and stimulate erythritol biosynthesis in cells and is not consumed by the yeast, therefore it remains in the broth after the end of the culture and has to be removed at the steps of product purification. However, the osmotic pressure increase might also be triggered by increasing the concentration of other medium components, including the substrate. Therefore, we compared the effect of osmotic pressure regulation by the increase of substrate concentration and NaCl in erythritol biosynthesis by Y. lipolytica ( Table 3). The initial osmotic pressure of the medium was increased to the level of about 3.0 Osm/kg by the simultaneous application of salt (32.5 g/L) and glycerol (150 g/L) or only by the substrate (220 g/L). In comparison to the process with NaCl, in the culture conducted without salt the erythritol productivity remained at a high level of 0.95 g/Lh, and higher availability of the substrate resulted in the production of 93.0 g/L of erythritol. The use of CSL did not have any negative impact on the activities of glycerol kinase, glycerol-3-phosphate dehydrogenase, or transketolase (Table 3), compared to the culture with Y. lipolytica K1 grown on media with NH 4 Cl being a mineral source of nitrogen [16]. The higher glycerol concentration slightly stimulated the activity of glycerol kinase, the enzyme responsible for the first step of glycerol consumption. Moreover, the increased osmotic pressure stimulated by glycerol and by the presence of both glycerol and NaCl enhanced the activities of transketolase and erythrose reductase, which correlated with the increased amount of erythritol produced by the yeast in these cultures and is in agreement with earlier studies [16,36].
The other contrivance for cost reduction is to eliminate the unnecessary ingredients of the medium. As mentioned above, CSL is a complex substance of natural origin, and might be used, not only as a source of nitrogen, but also as a source of vitamins and trace elements in a fermentation medium [37]. Thus, in the next experiment we have verified the possibility of eliminating mineral components of the medium and performed erythritol biosynthesis in the simple medium containing only the substrate and CSL (Table 4). When pure glycerol was applied, the production of erythritol and by-formation of mannitol were similar in the medium with and without the minerals; however, the elimination of minerals resulted in prolonged biosynthesis time, which was not a desired effect. This was not observed when crude glycerol was used as the substrate, as there were no significant differences between the culture with the rich medium and the medium without minerals. It is well-known that crude glycerol contains impurities of natural origin, e.g., vitamins and trace elements [38]. Therefore, it might be concluded that the combined application of CSL and crude glycerol in the performed experiment provided a sufficient level of nutrients and minerals for erythritol biosynthesis.
In the next step, erythritol production was also performed in the simple medium with CSL and without mineral supplementation; however, an amount of glycerol (20 g/L) was replaced by the equal portion of the CFFA. The results obtained showed that erythritol biosynthesis by Y. lipolytica was possible in the medium composed only from waste products such as CSL, CFFA, and crude glycerol, which offers a highly cost-effective possibility for the commercial process. Concentration of biomass (26.4 g/L) and erythritol (108.1 g/L) obtained in the two-steps biosynthesis process, in the culture with 200 g/L of crude glycerol and 20 g/L of CFFA (Table 5), were comparable to the concentrations achieved in the process performed on the low-cost medium with 220 g/L of crude glycerol ( Table 4).
Purification of the product from the culture broth obtained after the biosynthesis process might be the second cost-consuming factor of the production process. In the earlier study, we observed that Y. lipolytica Wratislavia K1 preferred utilization of by-products of erythritol biosynthesis (i.e., mannitol, arabitol, organic acids) rather than erythritol consumption from the culture broth [7,39]. The unique feature of this strain enables following the idea of cleaner production through the implementation of the natural purification process in which the selectivity of erythriol in the post-culture broth might be increased. In this study, the broth obtained after erythritol production by cultures with Y. lipolytica A-6 was collected, separated from biomass, sterilized, supplemented with nutrients, and used as the medium for the Wratislavia K1 bioreactor batch culture. For comparison, the process was conducted at pH 3.5 and 4.5, which are the conditions close to promoting erythritol [16] and biomass production [20], respectively. The comparison of the composition of the initial medium ( Figure 2) and the broth after cultivation showed that the selectivity of the erythritol increased from 72% to 97% in the culture performed at pH 3.5. At pH 4.5, its selectivity was slightly lower (96%); nevertheless, the process was significantly shorter, as it lasted only 49 h. It should be mentioned that, on an industrial scale, purification of erythritol from a mixture of polyols is problematic due to their similar chemical structures. Increasing the selectivity of erythritol in the culture broth is therefore beneficial, although further purification steps are required to obtain the final commercial product. After the process performed in this study, the biomass level reached 22 g/L, and its protein content reached 40.4%, meeting the standards recommended for fodder yeast. Worth noticing is that Y. lipolytica yeast is non-pathogenic and was recognized as safe (GRAS status) by the Food and Drug Administration (USA) and might be successfully applied for the fodder purposes [40].

Conclusions
It seems that one of the most important tasks for biotechnology is not so much the possibility of proposing new compounds and products but their production through environmentally-friendly processes. So far, the interest of scientists has focused mainly on the use of industrial waste as a substrate for the biosynthesis of new compounds. It should be remembered, however, that the proposed process will only be effectively attractive if it does not generate new waste and does not require the use of compounds adverse to the natural environment. In this work, Y. lipolytica A-6 was able to produce 108.0 g/L of erythritol with a production rate and yield of 1.04 g/Lh and 0.45 g/g, respectively, only in the two-component low-cost medium containing only crude glycerol and CSL. In turn, the use of Y. lipolytica Wratislavia K1 was proven to be effective for the natural purification of the culture broth and resulted in an increase of erythritol selectivity from 72% to 97% and a production of 22.0 g/L of biomass with a high protein content (40.4%). Summarizing, in this study, we have proposed a modern environment-friendly process by Y. lipolytica yeast that allows valorization of crude glycerol into erythritol, natural purification of the culture broth, and by-production of biomass suitable for animal feed.